Sahi-Detectron2

[Alisoltan82]

detection_model = AutoDetectionModel.from_pretrained(
    model_type='detectron2',
    model_path='./model_final (3).pth',
    config_path= "COCO-Detection/faster_rcnn_R_101_FPN_3x.yaml",
    confidence_threshold=0.1,
    device="cpu",
    category_mapping = {
        '0':'category 1',
        '1':'category 2'
    }
    
)

Here is the problem, loading detectron config file comes with its labels and categories, mapping or remapping doesn't work in this case and I tried both and deliver this error

KeyError                                  Traceback (most recent call last)
[<ipython-input-85-65128e870783>](https://localhost:8080/#) in <cell line: 1>()
----> 1 result = get_prediction('/content/00be38a5c0566291168fe381ba0028e6.png', detection_model)
      2 
      3 
      4 
      5 

3 frames
[/usr/local/lib/python3.10/dist-packages/sahi/models/detectron2.py](https://localhost:8080/#) in <listcomp>(.0)
    168                     segmentation=None,
    169                     category_id=category_id.item(),
--> 170                     category_name=self.category_mapping[str(category_id.item())],
    171                     shift_amount=shift_amount,
    172                     score=score.item(),

KeyError: '75'

advice in loading config file or mapping classes that my model trained on
Thanks for support

[fcakyon]

Have you checked this tutorial? @Alisoltan82

https://medium.com/codable/perform-sliced-inference-and-detailed-error-analysis-using-detectron2-models-37063e33073f